Formaldehyde electron density distribution

Hydrogen bonding must have an effect on the electron density distribution of a molecule. In principle, this should be observed in the deformation density distributions discussed in Chapter 3. There are, in fact, two methods available. One is purely theoretical, in which the calculated deformation density for a hydrogen-bond dimer or trimer is compared with that of the isolated molecule. The other method compares the experimental deformation density of a hydrogen-bonded molecule in a crystal structure with the theoretical deformation density of the isolated molecule. Formamide has been studied by both methods [298, 380], and there appear to be significant differences in the results which are not well accounted for. Theoretical difference (dimer vs. monomer) deformation density maps have been calculated for the water dimer and the formaldehyde-water complex [312]. When those for the water dimer are decomposed into the components described in Chapter 4, a small increase in the charges on the atoms in the O-H -O bond due to the charge-transfer component is predicted [312]. 

Similar arguments also apply, if one wants to compare the CNDO/2 electron densities of dimethylcarbodiimide (213) (99) with CNDO/S electron density distributions in other cumulenes. There cannot be observed any correlation (r > 0.85) between CNDO/2 (37a) and CNDO/S electron densities (tested for ketene (43), ethylene, formaldehyde (H2C==0), formic acid (HCOOH), and diazomethane (42) (101)). 

Figure 17.2 compares the electrostatic potential surfaces of ethylene and formaldehyde and vividly demonstrates how oxygen affects the electron distribution in formaldehyde. The electron density in both the o and tt components of the carbon-oxygen double bond is displaced toward oxygen. The carbonyl group is polarized so that carbon is partially positive and oxygen is partially negative. 

Using the combination of main-frame CDC 6400 and Tektronix computations, a number of phenomena were studied with electron density functions, and especially with projection plots. Particularly useful were plots of difference functions in which the electron distributions of isoelectronic systems were compared directly. In such applications, we noted that a corresponding difference plot of the electron density itself in any given plane is not meaningful since the number of electrons may change that is, from one compound to the next the electron density can shift from one plane to elsewhere. In the projection plot the total number of electrons remains the same for both species and the integral of an isoelectronic difference function must sum to zero. Some examples of the kinds of problems studied are the vm transition of formaldehyde, substituent effects in substituted benzenes, and polarization... 

---